EP96 CDPs in Action: Open-Source Solutions & Future of Customer Data Platforms
===

[00:00:00] 

[00:00:06] DARIN NEWBOLD: Welcome to Commerce Today. My name is Darin Newbold. And with me is always Josh Warren. We are here to talk about what's happening in the world of commerce today. And well, for today, we're continuing from last time talking about CDPs. And now we're looking really at open source solutions and, and what that future is for these customer data platforms.

[00:00:27] DARIN NEWBOLD: But Josh, maybe to bring us up to speed, share a little bit, kind of a recap from last

[00:00:32] JOSHUA WARREN: Yeah, so last week's episode, if you haven't checked it out, definitely go check it out. we covered customer data platforms, what they are, kind of the basics, even compared them and contrasted them against CRMs. And we talked through, five or six of the big commercial, CDPs.

[00:00:48] JOSHUA WARREN: So, kind of the, the enterprise grade, and then, Yeah. We saved the best possibly for last for this week's episode to really dive into open source CDPs. And I think a lot of people out there probably [00:01:00] don't even realize that there are open source CDPs. and there's some interesting options in that category.

[00:01:06] DARIN NEWBOLD: Well, I can't wait to hear more about it. And as you kind of helped me understand the difference between CRMs and, and CDPs and the, and the, just the vast amounts of data that they're having to look at.

[00:01:16] DARIN NEWBOLD: So as we begin this, I guess, tell us more, we've talked about an open source, but is it really a viable opportunity, alternative to these commercial CDPs? 

[00:01:27] JOSHUA WARREN: these commercial CDPs?

[00:01:38] JOSHUA WARREN: Then you can do as much or honestly more with an open source CDP than you can do with a commercial CDP. It is open source. You have access to the source code. There's a level of customization and modification that you can do that you just can't do with a commercial CDP. however, it does have that requirement of that technical expertise.

[00:01:58] JOSHUA WARREN: So I think a lot of people, [00:02:00] they hear open source and they gravitate towards it because they go, Oh, that's that free stuff. Well, it is free, but as they always like to remind us, open source more than anything is free as in speech, not necessarily free as in beer, so

[00:02:14] DARIN NEWBOLD: Important safety tip. You

[00:02:15] JOSHUA WARREN: have access to the code. If you make modification, depending on the license, if you make modifications, the code, you actually have to contribute it back.

[00:02:21] JOSHUA WARREN: So it creates this ecosystem and community of basically freely exchanging the technology. It doesn't mean you get a free CDP, though. You still need to host it. You still need to maintain it. So there are definitely still costs, but again, for the right company with a technical team, either in house or at their agency, those costs often are substantially lower than the enterprise CDPs.

[00:02:44] DARIN NEWBOLD: Oh, I would think that would be, yeah, significantly lower when you can, if you have the knowledge to make the, make the changes yourself. I love the example of free speech, not free beer, because I've run into that when I've gone down the path of open source [00:03:00] hoping for some free beer and, you know, I was very disappointed.

[00:03:03] DARIN NEWBOLD: All that aside, let's let's kind of start going through some of those top ones that, that you really liked as far as open source CD, CDPs. Which ones do you see as kind of ones that you really like and, and the things you like about

[00:03:16] JOSHUA WARREN: Yeah. Yeah. So I wanted to highlight three. There's actually more, which again, kind of surprised me just how big this category is getting and how many open source players there are in it.

[00:03:24] JOSHUA WARREN: But we're going to look at three and that's Rudderstack, Apache Unomi, and PimCore CDP. So I'll start out with Rudderstack. It's very much, it's, I mentioned to you before that CDPs are all about storing lots of data. So you might even think, Oh, they're kind of a specialized data warehouse. That's Rutterstack's approach.

[00:03:44] JOSHUA WARREN: It's very much a data warehouse native and warehouse first approach. It's also privacy first. There's a lot of, a lot of GDR, GDPR compliance that can come up with CDPs. Gotta throw more acronyms in there as always. [00:04:00] and, Rutterstack really covers those well. and also, Just has all of these have a level of flexibility, but rudderstack is really designed for a much higher level of flexibility and rudderstack and PimCore both come in an open source version and in a paid version, so it gives you some options there, especially on the open source side, though.

[00:04:21] JOSHUA WARREN: I feel like rudderstack probably requires the second most. Technical resources of these three. I think the Apache solution requires the most. but if you want to store your customer data in your own data warehouse, where you have complete and total control over it, you own everything from the code to the servers.

[00:04:41] JOSHUA WARREN: especially if you're in Europe and you're in a situation where you're like, we just don't want to have partners involved in this. We want it all to be in house. the rudder stack open source solution fits really well for that. That's

[00:04:52] DARIN NEWBOLD: That's awesome. Well, I guess let's go into that next one. The Apache, you know, me, or you don't know

[00:04:58] JOSHUA WARREN: know me.[00:05:00] 

[00:05:00] JOSHUA WARREN: and it's interesting because there's, there's a little bit of, Controversy. There's some people out there that say, Oh, you know, he's not really a full CDP. It's not there yet. it's been around long enough and the Apache project, they call it a CDP. It is a CDP. it may not have all the same features as some of the other ones, but it is GDPR compliant.

[00:05:23] JOSHUA WARREN: and it is very highly customizable. and something that I didn't mention Apache.

[00:05:27] JOSHUA WARREN: You may, if you're a, an e commerce geek that likes to deal with servers, you've heard of Apache before, but the rest of y'all, Apache is the same group that produces the Apache web server, which is, one of the most commonly used open source web servers on the internet.

[00:05:45] JOSHUA WARREN: It powers a substantial portion of the internet. So there's already a lot of companies that are using one open source Apache project and are kind of used to. Being in that ecosystem and that's where Apache, you know, me really fits is if [00:06:00] you're all in on Apache, then it just makes sense to use, you know, me, however, while it's been around a while, there's that debate over.

[00:06:06] JOSHUA WARREN: Is it really a CDP or not? so it has. Kind of a less mature ecosystem and hasn't really had the ecosystem form around it that Pimcorn, Rudderstack has, or even that like the Apache web server has. you also need more, development resources for this one than any of the, open source CDPs we're going to talk about.

[00:06:26] JOSHUA WARREN: But if you want a very high level of customization and you're already kind of bought in and using the Apache world of, open source projects, then Apache Unomi makes sense.

[00:06:36] DARIN NEWBOLD: Interesting.

[00:06:37] DARIN NEWBOLD: A question for you, Josh, before we jump into PimCore, just do these, CDPs, do they all, these open source ones, do they all use the same tech stack as far as from programming language and those

[00:06:49] JOSHUA WARREN: they, they don't. And that is, that is a good call out and a good thing to look at is like, for instance, Rutterstack, Their one big dependency is a database server called Postgres [00:07:00] SQL that, without getting into a lot of history around open source databases, I'll just say that Postgres is what I learned back in 1998.

[00:07:09] JOSHUA WARREN: haven't used it a lot since then. There are definitely niches and projects and areas of data science where it's used heavily. But like in the e commerce world, it's not our go to default database. So

[00:07:21] DARIN NEWBOLD: Okay. And then what about Apache?

[00:07:23] JOSHUA WARREN: I'm actually not sure, what it uses. I haven't dug into it as much as I've

[00:07:27] DARIN NEWBOLD: I'm almost wondering if it's uses, at least from a programming PHP, I think, because doesn't the Apache server, isn't that

[00:07:35] JOSHUA WARREN: web server itself is C or C plus plus.

[00:07:39] JOSHUA WARREN: so

[00:07:40] DARIN NEWBOLD: I'm wrong. Anyway,

[00:07:41] JOSHUA WARREN: good question though.

[00:07:42] DARIN NEWBOLD: no worries. No worries. Maybe we'll find that for next time, or if you already know, definitely let us know. We'll, we'd be happy to

[00:07:47] JOSHUA WARREN: well, and you've got to already got me thinking about, some of our audience may have noticed. I like turning these episodes into a LinkedIn newsletter every week.

[00:07:55] JOSHUA WARREN: So in the newsletter, I will go back and actually find what's the underlying [00:08:00] program in the language and database for each of these three. And I'll include that in the newsletter. So

[00:08:04] DARIN NEWBOLD: Well, let's get with it here. PimCore, we've teased it. Let's talk about it because it looks like it's one of your, probably one of your favorites, I would

[00:08:11] JOSHUA WARREN: It is. and a lot of that's because if you've heard of Pymcor, it's a Pym, but it's in the name, product information manager, but Pymcor actually has a whole suite of products. They have the Pym, they have a DAM, digital asset management solution, they have content management system, and they have the CDP.

[00:08:29] JOSHUA WARREN: So, and with all of these products, they have an open source version and then they have a paid kind of enterprise version. So. You can build out a entire solution based on everything but your core e commerce platform using nothing but PimCore technologies. and so it gives you this ability to flex and kind of take that composable commerce approach and bring in pieces here and there.

[00:08:52] JOSHUA WARREN: Knowing that you can also then scale up to their full suite of solutions. You can scale into their enterprise offering. it is like last [00:09:00] week we talked about how, if you just wanted to go launch a CDP, just go sign up for Twilio segment because for a relatively small amount of money, you can be up and running pretty fast.

[00:09:09] JOSHUA WARREN: PimCore CDP, you can't just go launch the open source PimCore CDP and be up and running. so it is in some ways a little bit of overkill for the smallest businesses, but I do think that, It's a good fit for most. you will need a technical team for implementation. not nearly as much as you need for Apache.

[00:09:28] JOSHUA WARREN: You know me. So if you're already using PimCore suite of tools, or if you're looking for more than just a CDP, I definitely would look at the open source version of PimCore

[00:09:39] DARIN NEWBOLD: like that would be a kind of a no brainer to make that decision. Well, You talked about a little bit last time, but now that we've talked these three CDPs, kind of how would you compare them to their commercial competitors, if you will?

[00:09:54] JOSHUA WARREN: yeah, so you save on licensing fees. There are no licensing fees. you do [00:10:00] then have to pay hosting fees, and then development fees. so it's really, it gets a little tricky to calculate out a comparison between the two.

[00:10:09] JOSHUA WARREN: without really diving into your specific situation. If you already have a technical team, if you already. Have some servers that you're running, then the cost can be a lot lower. if you are going to have to spin up a new technical team, you're going to have to spin up new servers. You've no one in your company knows anything about hosting.

[00:10:28] JOSHUA WARREN: then again, it's free as in speech, not free as in beer. however, what you get for that, that little bit higher overhead around the technical expertise that's required, you get full control. something I always like to point out is. Whether you're working with a SaaS or just really any closed platform, if they decide to remove a feature, you're out of luck.

[00:10:52] DARIN NEWBOLD: Yep.

[00:10:52] JOSHUA WARREN: Open source platforms, you have total control. If something gets removed, you can just add it back in. You have the source code. You [00:11:00] can do what you want to do with it. It gives you a level of control. It also gives you transparency. in this world of everyone Taking our data and using it to train their own proprietary AI's because I got to mention AI every week

[00:11:12] JOSHUA WARREN: with an open source solution.

[00:11:14] JOSHUA WARREN: You can look through the code. You say, okay, nothing's phoning home and sending our data out anywhere else. but then also, especially with a project like Apache, you know, me, It's community driven, so it doesn't always have the regular release cadence that a commercial solution would have. this isn't as big of a risk with PimCore or Rudderstack, just because they're backed by a commercial company.

[00:11:36] JOSHUA WARREN: But lots of times with open source projects, you'll see it kind of ebb and flow of the community gets really fired up, they release a lot of updates, and then maybe nothing for months or even years.

[00:11:45] DARIN NEWBOLD: out. The community support

[00:11:49] DARIN NEWBOLD: can be so great because you can have some great, great knowledge, but sometimes it can be a little challenging to find, to dig to that. So you're going to have, you may have to search a little bit. [00:12:00] Well, And I already kind of, I guess I'm teasing it a little bit as we go into it.

[00:12:04] DARIN NEWBOLD: What are, what are some of the challenges that, that you run into when you're implementing a open source CDP?

[00:12:09] JOSHUA WARREN: Well, actually these challenges, Really, most of these challenges will be for both

[00:12:15] DARIN NEWBOLD: both of them. Okay. Commercials. 

[00:12:17] JOSHUA WARREN: so with both you have data privacy concerns.

[00:12:19] JOSHUA WARREN: So we've already mentioned GDPR. There's also CCPA, which is the

[00:12:25] JOSHUA WARREN: California Consumer Privacy Act. There's about 10 other states working on their own version of the CCPA. So now we're going to have 10 more acronyms, but basically there's a lot of rules out there about what you do with customers information.

[00:12:38] JOSHUA WARREN: even the line that is drawn of what becomes information you really have to treat with extra care. used to it was like, you know, don't lose their credit card number. Now even, certain, certain information and behavioral data that some companies may think, Oh, that's not that big of a deal. Some of these laws now say that has to be protected just as much as say their financial [00:13:00] information.

[00:13:00] JOSHUA WARREN: CDPs can help though, because they centralize your efforts around storing that customer data. So if you select a CDP that is, GDPR and CCPA aware, it will already have features built in, like in the GDPR, there's a right to forget. So basically a customer can say, I want you to lose all of the data you have about me and wipe it from all of your systems.

[00:13:23] JOSHUA WARREN: And those, CDPs that have support for that, that's just a button you click. So

[00:13:28] DARIN NEWBOLD: That is interesting.

[00:13:29] DARIN NEWBOLD: Well, what about, what about things around the integration? And, and obviously costs are different. We've, we've teased that a little bit, but tell us a little bit more there.

[00:13:38] JOSHUA WARREN: about that. Yeah, so around the integrations, integrating a CDP into your existing systems, especially older systems.

[00:13:45] JOSHUA WARREN: There's a lot of systems out there that were designed before CDPs were a thing, and so they don't really have pre built integrations. there can be a lot of, custom work that has to be done around those integrations. this is where. Especially the [00:14:00] commercial solutions we mentioned last week. If you're using Adobe products, the Adobe CDP is pre integrated.

[00:14:05] JOSHUA WARREN: If you're using Salesforce products, their CDP is pre integrated. and then on the open source side, I would really look at, both PimCore and Rudderstack because they have more, Prebuilt integrations. but if you venture out into one of those other solutions that doesn't have as many prebuilt integrations, or if you have a lot of custom, applications running within your business, then you're going to have some more challenges.

[00:14:28] JOSHUA WARREN: And that's another one where it really, you got to take it on a business by business basis and look at, okay, what, what API methods does that software support? How are we going to integrate with it? yeah. It can be done, but it's a project for sure.

[00:14:42] DARIN NEWBOLD: sure. Your mileage may

[00:14:43] JOSHUA WARREN: Exactly. And then you mentioned the costs and yeah, I think this is going to be a whole separate article as well.

[00:14:50] JOSHUA WARREN: because again, with the open source CDPs, you have no license fees, but you might have higher maintenance fees. with some of these CDPs that are pre integrated versus ones that aren't, you [00:15:00] might have different integration costs. So again, you really have to, Look at the total cost of ownership and you have to have someone calculate it for your exact needs and not just, Hey, someone told me this one has the lowest total cost of ownership because realistically every platform, their salesperson will say they have the lowest TCO and the highest ROI and you've really got to calculate that out for your business.

[00:15:25] DARIN NEWBOLD: Absolutely.

[00:15:25] DARIN NEWBOLD: Well, let's talk, as we're, as we're kind of getting on the downhill side of this, let's talk the future of CDPs and e commerce and, and what that's looking like. And your favorite topic is about to come

[00:15:36] JOSHUA WARREN: Oh, yeah. So AI driven personalization. So I was

[00:15:41] JOSHUA WARREN: Before the episode about how CDP versus CRM CDP has so much more behavioral data like it typically is storing so much more really minutia that a person can't really go through and say, okay, at this time, they clicked this from this device.

[00:15:55] JOSHUA WARREN: They did that. They did that. Now I think we should show them this offer. An AI [00:16:00] can do that though. And so there are CDPs that are being used in really smart ways to do AI driven real time personalization. and you can really just hyper personalize the customer experience. And so, something that, Amazon has been doing more of is predictive analytics for figuring out what you're going to buy before you even realize you want to buy it.

[00:16:22] JOSHUA WARREN: and they take it a level further than that. Not only do they use that on a micro level to target you and your experience and show you the things that they think you're going to want to buy on a macro level, they will say, okay, we think people in Dallas, Texas are about to start buying a lot more of this particular product.

[00:16:39] JOSHUA WARREN: We're going to shift more of it into our fulfillment centers in Dallas so that when that demand hits, we can deliver it same day.

[00:16:45] DARIN NEWBOLD: Wow. That's crazy. Well, they also send it to their Amazon lady devices that basically tell me I need to order stuff when I'm not ready to order stuff.

[00:16:55] JOSHUA WARREN: Oh yeah. Oh, they love to send those reminders of don't you think it's time to order one of these?

[00:16:59] JOSHUA WARREN: [00:17:00] So, yeah, yeah. but also, in the future, I think we're going to see even more real time. I mentioned the real time personalization, but just real time data processing in general. I mean, these CDPs are especially the kind of leading edge ones. Are getting to a point where, say, on Black Friday, you're going to be able to see, okay, this is what our consumer behavior or customer behavior is doing this versus this is what we expected it to do.

[00:17:26] JOSHUA WARREN: Do we need to make some shifts in our marketing campaigns this afternoon, even to make sure we reach our targets for Black Friday?

[00:17:34] DARIN NEWBOLD: That's crazy. Privacy, data protection, as well as integrations. So

[00:17:38] JOSHUA WARREN: even more laws, as I mentioned, more data privacy laws are coming out by centralizing and using a CDP. You can centralize kind of your compliance with those laws into that system.

[00:17:48] JOSHUA WARREN: And again, the better CDPs are staying ahead of those laws and making sure they have the features built in to support them. And then I'm really excited about CDPs, in the world of composable [00:18:00] commerce, especially if you have a PIM. Product information manager, you have a CMS content management system and you have a CDP customer data platform.

[00:18:09] JOSHUA WARREN: And all of these are integrated in a very composable commerce like way. Then changing your e commerce platform becomes so much easier because you're no longer having to migrate all this data. You just. change from one integration to another. and that's where, we've talked before about the platform 0 was released just last week, and there's already some interesting things I'm seeing between, say, the PimCore products and Medusa, and just what that makes possible.

[00:18:35] JOSHUA WARREN: And I think we're going to see a lot more of that in the composable world.

[00:18:39] DARIN NEWBOLD: That is fantastic. Well, well, listeners and, and everyone that's, that's taking part in this, gosh, I really hope you've gotten a great overview of CDPs, both, in our previous episode, the commercial ones, and now in this episode, the open source, the, the good, the bad, the uglies, yes, your mileage will vary.

[00:18:59] DARIN NEWBOLD: And, [00:19:00] and as Josh always tells us, it really depends on your situation and you've got to evaluate it because there's no one size. Fits all. Last thoughts here on these open source CDPs, Josh.

[00:19:11] JOSHUA WARREN: So I will have an article out on my LinkedIn, talking more about the open source CDPs and I would invite you while you're there. you'll find me on LinkedIn as Joshua Warren with a Creatuity Gold background in my headshot and, just reach out to me and tell me kind of what you're, you're seeing in your own, content.

[00:19:29] JOSHUA WARREN: Work around CDPs, and if you're interested, I still have those, 30 minute free e commerce problem solving sessions. Would love to talk through your specific use case of CDP.

[00:19:40] JOSHUA WARREN: opportunities 

[00:19:41] DARIN NEWBOLD: learn more. Well, as always, we appreciate you being here. Please like, subscribe, follow all of the groovy stuff. And as always, take care and we wish you all the best here at Commerce Today. Take care.

​ [00:20:00]